Using accommodation price determinants to segment tourist areas

Abstract

Accommodation services oriented to different tourist segments usually have different price determinants. Thus, in multi-facet destinations such as large regions or cities, it should be possible to find and describe the underlying types of tourism in the destination by using a price determinant analysis. In this paper, a methodology based on stepwise geographically weighted regression (GWR) is developed, using a k-means clustering algorithm to determine the different types of tourism existing in a large geographical area. The method is applied to the island of Gran Canaria (Canary Islands, Spain), using a database of more than 2000 peer-to-peer accommodation units spread over the geography of the island. As a result, it was possible to identify and classify eight different clusters of types of tourism within this geographical area. This methodology can be used in other geographical areas to identify the different types of tourism developed in them.